A new exchange-traded fund providing exposure to extreme ultraviolet lithography companies has doubled in price since its launch in April 2024 [4].
This growth reflects the critical role of EUV technology in producing the advanced semiconductors that power artificial intelligence. Because EUV lithography is the only method capable of producing sub-five nm nodes, it serves as the foundational layer for AI accelerators like Nvidia’s H100 [1, 2].
The EUV ETF tracks semiconductor equipment manufacturers, with a primary focus on ASML Holding NV [1]. Based in Veldhoven, the Netherlands, ASML is the sole provider of the machines required for this process [1, 3]. The company's latest High-NA EUV machines carry a price tag of approximately $400 million each [5].
To meet the surging demand for AI chips, ASML planned to build at least 60 EUV machines in 2024 [6]. This expansion comes as U.S. hyperscalers prepare for massive infrastructure investments, with plans to spend almost $700 billion in capital expenditures in 2026 [7].
Despite the increase in production, industry analysts disagree on whether the supply chain can keep pace. Some reports suggest that a critical EUV bottleneck continues to hold back the broader AI boom [2, 7]. Others point to the aggressive manufacturing targets set by ASML as evidence of a rapidly expanding capacity [6].
The ETF trades on U.S. exchanges, allowing investors to bet on the hardware layer of the AI revolution without holding individual semiconductor stocks [1]. This financial instrument highlights the shift in investor interest from software applications to the physical machinery that makes those applications possible.
“The EUV-focused ETF has doubled in price since its launch last month.”
The rapid ascent of the EUV ETF underscores a market realization that AI scalability is not just a software or design challenge, but a physical manufacturing constraint. By concentrating on the lithography bottleneck, investors are hedging against the volatility of chip designers and instead betting on the indispensable infrastructure provided by ASML. If production cannot scale to meet the $700 billion projected spend by hyperscalers, the hardware supply chain will remain the primary limiting factor for AI growth.
